To understand the mechanism for the development of heavy snowfall and the influence of the variation in the satellite-based Sea Surface Temperature (SST) distribution on the evolution of snow convection cells, synoptic characteristics of heavy snowfall over the Honam District, south-western Korean Peninsula, were analysed during December 2005; several numerical experiments were also conducted. New Generation Sea Surface Temperature (NGSST) data based on satellite observation are used in this study, with Penn State University/National Center for Atmospheric Research (PSU/NCAR) 5th- generation Mesoscale Model (MM5) used as the numerical model. Since the cold air from the arctic centre spreads strongly along the planetary wave, the Siberian high pressure matured earlier than normal in 2005. The analysis of the Arctic Oscillation Index (AOI) also indicated the development of a cold Siberian high. Snow convection cells occurred primarily as a result of air-sea interactions and orographic forcing over the Korean Peninsula and, in this case, two major convection cells appeared on 21 December 2005. Because of the decrease in the SST with time, the intensity of the convection cells also decreased with time. The two-dimensional distribution of the SST was also strongly associated with the amount of inland snowfall. Therefore, satellite-based observation is a useful method to detect the change in detail of SST distribution, and the exact estimation of the SST gradient in the central Yellow Sea and near the coast is also an important factor in forecasting the intensity of snowfall over the south-western Korean Peninsula.